In studying negative regulation of hematopoietic cell growth, we have found that transforming growth factor beta is a potent inhibitor. TGF-beta1 and TGF-beta2 are equipotent inhibitors which have a selective effect on hematopoiesis by halting the growth of only early human and murine progenitors. By using immature murine hematopoietic cells that respond to IL 3, it was shown that TGF- beta acted directly on the cells to block multipotent cell development but not unipotent cell growth. The earliest measurable events in vitro such as high proliferative potential colony formation (HPP-CFU) and induction of Thy-l antigen were also inhibited by TGF-beta. In contrast, there was no effect on G-CSF stimulation of granulopoiesis. In euthyroid development, CFU-GEMM and BFU-E but not CFU-E were inhibited. Thus, the ability of TGF- B to block hematopoietic progenitor cell growth depends on the differentiated state of the cell and suggests that TGF-beta is an important regulator of hematopoiesis. Fresh AML and CML as well as early myeloid leukemic cell lines retained sensitivity to positive stimulation of CSFs and negative inhibition by TGF-beta. The lone exception was HL-60, a human promyelocytic cell line, which was found not to have receptors for TGF-beta. However, agents that stimulate differentiation of HL-60 caused a reappearance of the TGF-beta receptors on the cell surface. These results suggest that the unresponsiveness to agents can be overcome in leukemic cells. In contrast to myeloid leukemic cells, cultured lymphoid leukemic cells were insensitive to TGF-S inhibition, whereas normal lymphoid cells are sensitive to TGF-B. Thus, this loss of negative regulation may play a role in the growth of some leukemic and malignant lymphoid cells.